Modelling of early flame kernel growth towards a better understanding of cyclic combustion variability in SI engines

Highlights: • The cyclic variability in combustion was studied experimentally and numerically. • Newly developed quasi-dimensional ignition model (QDIM) is applied. • Quantitative picture about CCV causes at different operating conditions is given. • Variation of turbulence affects the cyclic combustion variability up to 66%. • CCV of early flame kernel growth are responsible for up to 18% of total CCV. - Abstract: The analysis of particular effects that influence cyclic combustion variability in spark-ignition engines is presented in this paper. The cycle-simulations are performed over 16 knock-free full load operating conditions of a single cylinder engine fuelled with gasoline. For the modelling of early flame kernel growth a newly developed quasi-dimensional ignition model is applied. The ignition model includes a detailed description of the electric circuit, the electric spark length, the spark plug geometry and the flame kernel growth. The main turbulent combustion is calculated using the extended quasi-dimensional fractal combustion model requiring the correct prediction of in-cylinder turbulence level. The cyclic combustion variability is simulated by a variation of the in-cylinder turbulence level, the flow angle at spark plug and the stratification of air equivalence ratio from cycle-to-cycle. The statistical analysis of the indicated mean effective pressure (IMEP) gives a more detailed quantitative picture of the causes of the cyclic combustion variability in SI engines over the different operating conditions. The variation of in-cylinder turbulence is found to be the dominant factor that affects the cyclic variability (up to 66%), while the variation in the early flame kernel growth rate is responsible for up to 18% of the overall cyclic combustion variability in the modelled SI engine at full load conditions